INDEX

Explanations

discussion thread

vocabulary lists and types

New Auto-Interp

Configuration

Prompts (Dashboard)

392,802 prompts, 256 tokens each

Dataset (Dashboard)

monology/pile-uncopyrighted

Embeds

IFrame

Link

Not in Any Lists

Negative Logits

 delicacy

0.52

ವರ್

0.48

 délic

0.47

 delicate

0.46

MULTIP

0.43

 capacitor

0.43

Agregar

0.43

引入

0.42

 crucible

0.42

 Manufactured

0.42

POSITIVE LOGITS

 เป้า

0.48

 त्यानुसार

0.47

decine

0.45

ละคร

0.42

ähl

0.42

 ይችላሉ

0.41

 எனப்படும்

0.41

dyl

0.40

 ชื่อ

0.40

zież

0.40

Activations Density 0.001%